Music Sequence

Music sequence research focuses on understanding and generating musical patterns using computational methods. Current efforts concentrate on developing generative models, such as variational autoencoders (VAEs) and diffusion models, often incorporating techniques from natural language processing like N-grams, to create realistic and emotionally expressive music, including multi-track and polyphonic compositions. This work is advancing both our understanding of musical structure and perception, as evidenced by studies correlating physiological data with musical improvisation, and enabling new tools for music creation and editing, such as instruction-guided remixing frameworks.